Text-to-Image Generation on T2I-CompBench (Perceptual Alignment)
0.799DINO ScoreOurs (DINO)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Ours (DINO)Backbone=SDXL-Turbo2025.12 | 0.799 | 0.424 | 0.764 | 0.351 | |
| Ours (DINO)Backbone=SANA-Sprint-1.6B2025.12 | 0.738 | 0.437 | 0.767 | 0.33 | |
| Ours (DINO)Backbone=PixArt-α2025.12 | 0.716 | 0.331 | 0.674 | 0.335 | |
| Parmar et al.Backbone=SDXL-Turbo2025.12 | 0.701 | 0.329 | 0.68 | 0.334 | |
| Parmar et al.Backbone=SANA-Sprint-1.6B2025.12 | 0.684 | 0.331 | 0.718 | 0.338 | |
| Parmar et al.Backbone=PixArt-α2025.12 | 0.59 | 0.256 | 0.593 | 0.328 | |
| i.i.d.Backbone=SDXL-Turbo2025.12 | 0.586 | 0.244 | 0.634 | 0.332 | |
| i.i.d.Backbone=SANA-Sprint-1.6B2025.12 | 0.562 | 0.252 | 0.656 | 0.334 | |
| i.i.d.Backbone=PixArt-α2025.12 | 0.469 | 0.188 | 0.512 | 0.326 |